Objective To investigate the performance of loading naringin composite scaffolds and its effects on repair of osteochondral defects. Methods The loading naringin and unloading naringin sustained release microspheres were prepared by W/O/W method; with the materials of the attpulgite and the collagen type I, the loading naringin, unloading naringin, and loading transforming growth factor β1 (TGF-β1) osteochondral composite scaffolds were constructed respectively by " 3 layers sandwich method”. The effect of sustained-release of loading naringin microspheres, the morphology of the composite scaffolds, and the biocompatibility were evaluated respectively by releasingin vitro, scanning electron microscope, and cell counting kit 8. Forty Japanese white rabbits were randomly divided into groups A, B, C, and D, 10 rabbits each group. After a osteochondral defect of 4.5 mm in diameter and 4 mm in depth was made in the intercondylar fossa of two femurs. Defect was not repaired in group A (blank control), and defect was repaired with unloading naringin composite scaffolds (negative control group), loading naringin composite scaffolds (experimental group), and loading TGF-β1 composite scaffolds (positive control group) in groups B, C, and D respectively. At 3 and 6 months after repair, the intercondylar fossa was harvested for the general, HE staining, and toluidine blue staining to observe the repair effect. Western blot was used to detect the expression of collagen type II in the new cartilage. Results Loading naringin microspheres had good effect of sustained-release; the osteochondral composite scaffolds had good porosity; the cell proliferation rate on loading naringin composite scaffold was increased significantly when compared with unloading naringin scaffold (P<0.05). General observation revealed that defect range of groups C and D was reduced significantly when compared with groups A and B at 3 months after repair; at 6 months after repair, defects of group C were covered by new cartilage, and new cartilage well integrated with the adjacent cartilage in group D. The results of histological staining revealed that defects were filled with a small amount of fibrous tissue in groups A and B, and a small amount of new cartilage in groups C and D at 3 months after repair; new cartilage of groups C and D was similar to normal cartilage, but defects were filled with a large amount of fibrous tissue in groups A and B at 6 months after repair. The expression of collagen type II in groups C and D was significantly higher than that in groups A and B (P<0.05), but no significant difference was found between groups C and D (P>0.05). Conclusion Loading naringin composite scaffolds have good biocompatibility and effect in repair of rabbit articular osteochondral defects.